Hi,
I am starting to uses Micress for the solidification of Silicon. I have managed to get some results using temperature coupling and grain growth competition but I am struggling to get a concentration simulation to work. I would like to model the silicon growth from a seed with nucleation of a second phase when the melt is enriched. My problem might be related to the phase diagram with very low partition coefficients (~1.E-4). I do not have a thermodynamic database for that material. I put the phase diagram in attachment.
How should I model that phase diagram? (FeSi2 phase can be considered stoichiometric)
Thanks a lot for your help.
Sylvain
Silicon solidification
Silicon solidification
- Attachments
-
- fesi.png (84.38 KiB) Viewed 1618 times
Re: Silicon solidification
Dear Sylvain,
Welcome to the MICRESS forum!
As far as I can see there should be no principal problem of modelling this system in MICRESS using a linearized phase diagram. As I understood you want to model growth of (nearly) pure Silicon starting from the melt, where later FeSi2 is expected to nucleate. FeSi2 would be defined as stoichiometric, and solid silicon can either be defined with solubility of Fe (small partition coefficient) or even as stoichiometric (pure Si) which would not make much of a difference. If you give a solubility to solid silicon, the phMin parameter (at the very end of the input file) should be as small as the partition coefficient, otherwise Fe would be trapped to a higher extent leading to a too high Fe composition in silicon.
Which are the problems you are struggling with? Did you already try to build up a corresponding input file? Can you describe the problems which occur and attach your input file?
Best regards
Bernd
Welcome to the MICRESS forum!
As far as I can see there should be no principal problem of modelling this system in MICRESS using a linearized phase diagram. As I understood you want to model growth of (nearly) pure Silicon starting from the melt, where later FeSi2 is expected to nucleate. FeSi2 would be defined as stoichiometric, and solid silicon can either be defined with solubility of Fe (small partition coefficient) or even as stoichiometric (pure Si) which would not make much of a difference. If you give a solubility to solid silicon, the phMin parameter (at the very end of the input file) should be as small as the partition coefficient, otherwise Fe would be trapped to a higher extent leading to a too high Fe composition in silicon.
Which are the problems you are struggling with? Did you already try to build up a corresponding input file? Can you describe the problems which occur and attach your input file?
Best regards
Bernd